BACKGROUND: Worms may protect against allergy. Early-life worm exposure may be critical, but this has not been fully investigated. OBJECTIVES: To investigate whether worms in pregnancy and in early childhood are associated with childhood eczema incidence. METHODS: The Entebbe Mother and Baby Study, an anthelminthic treatment trial, enrolled pregnant women between 2003 and 2005 in Uganda. Mothers were investigated for worms during pregnancy and children annually. Eczema was doctor-diagnosed from birth to age five years. A planned observational analysis was conducted within the trial cohort to investigate associations between worms and eczema. RESULTS: Data for 2345 live-born children were analysed. Hookworm was the most prevalent maternal worm (45%). Childhood worms were less prevalent. Eczema incidence was 4.68/100 person-years. Maternal hookworm was associated with reduced eczema incidence [adjusted hazard ratio (95% confidence interval), p-value: 0.71(0.51-0.99), 0.04] and modified effects of known risk factors for eczema: Dermatophagoides-specific IgE in children was positively associated with eczema incidence if the mother had no hookworm [2.72(1.11-6.63), 0.03], but not if the mother had hookworm [0.41(0.10-1.69), 0.22], interaction p-value = 0.03. Similar interactions were seen for maternal history of eczema {[2.87(1.31-6.27, 0.008) vs. [0.73(0.23-2.30), 0.60], interaction p-value = 0.05}, female gender {[1.82(1.22-2.73), 0.004 vs. [0.96(0.60-1.53), 0.87], interaction p-value = 0.04} and allergen-specific IgE. Childhood Trichuris trichiura and hookworm were inversely associated with eczema. CONCLUSIONS: Maternal hookworm modifies effects of known risk factors for eczema. Mechanisms by which early-life worm exposures influence allergy need investigation. Worms or worm products, and intervention during pregnancy have potential for primary prevention of allergy.